Comwerks Interactive

Blog Bites

 

Archive for the 'Measurement' category

Social Government Strategies (Part I): Using Blogs to Improve Customer Service

Posted by Gul Khan in Measurement, Social media on October 17th, 2008

Govgab: Hispanis Heritage MonthRecently we’ve been conducting some research that revolves around how the government can use new media and social media tools to help make their online offerings more meaningful to the people.

Address the Issues

We asked Singaporeans how the government could better serve the people from the online perspective and here is a summary of what they said:

1.    First, please get the basics right!
2.    The website it not relevant to me.
3.    I want to know more about you and the people behind the organization
4.    I want to connect with more people like me.
5.    I want to get involved and play a part.

“How to” Blogs

From the above, we can see that there definitely is scope for social and new media tools to solve all of the above and we have come up with a range of different strategies. And if you look at points (4) and (5) above, we can see that there are real communities connected to each government agency and their websites can play a key role in fostering and engaging in a conversations with these communities.

For now, I would like to explore a very simple and very workable strategy and that is to use blogs to improve service delivery and usage of online government services. This can solve points (1), (2) and in a limited way, (3). I guess you can call them “how to” blogs.

Gov Gab: The U.S. Government Blog

A great example of something along these lines would be Gov Gab blog. Here’s an excerpt of what they are about:

“In our daily jobs, we encounter a staggering amount of U.S. government information and services that can benefit your life. From saving money and visiting National Parks to finding out about government auctions and the latest recalls, we want to bring these resources to you in a new way—through our blog.”
Source: http://blog.usa.gov/roller/govgab/page/bios

As far as I understand it (please correct me if I am wrong), their blog helps people around the government maze to help deal with their day-to-day issues. The example below gives tips about how to “bank safely” during these troubled times:
•    It talks about and links to the 2008 Stabilization Act
•    It advises people on what to do when a bank fails (linking to the FDIC website)
•    It provides tips on how to choose a new bank (linking to the consumeraction.gov website)

Govgab: Banking Safely

Content Strategy – Anything That Is Current!

There’s one constant that almost every government website needs to be able to cope with: There is always something new to handle! New issues, requests, problem, question, can come up because of economic changes, policy updates, seasonal factors (e.g. tax filing month), things happening in another part of the world, changes in tastes, technology, health scares (e.g. SARs)…you name it.

So there is great scope for new content (aka new and relevant blog posts). If you belong to a government agency, just check out the current hot items in your:
•    Customer support logs
•    Website keyword searches
•    Customer feedback data, surveys etc

For example, from our own continuous learning programs, we’re noticing a LOT of questions coming in about the “enhanced baby bonus”. Information about this topic alone can span different government agency websites. As long as there is someone in the government keeping their finger on the pulse, a typical blog post about the baby bonus can be quickly created and it can link to the different websites that deal with this same issue but in a different way.

I think it’s an approach worth exploring for the Singapore government.

 
 

Search remains King..but is that the complete picture?

Posted by Keith Tan in Marketing, Measurement on September 19th, 2008

According to research firm eMarketer, search-ad spending will reach $10.4 billion this year, doubling what will be spent on display ads.

Online Advertising Spending (US)This means that Google, who currently control more than 70% of the U.S search market, would retain its dominancy, as search represents 42% of all ad spending.  Display ads currently account for half that figure.

So yes. This must obviously mean that those plain text ads are the safest and most cost effective way of spending your marketing budget.

Well…not really. In fact, search and display ads do not have to be mutually exclusive.

Atlas Solutions’ latest report findings showed that “When marketers supplement search with display impressions, they get a significant lift in conversions. Unfortunately, most advertisers that run both search and display are unaware of this…” The study demonstrated that “users exposed to both search and display ads convert at a higher rate: an average of 22 percent better than search alone and 400 percent better than display only.”

Now did those stats make you do a double take? But hey why not, it does make sense.

Display ads are great for driving brand awareness while search ads, being contextual in nature, have always provided better clickthroughs and conversions. Combine them both and you’ll stand a higher chance to closing that sale.

“The study, entitled “Close the Loop: Understanding Search and Display Synergy,” found that when combined, search and display advertising deliver profoundly better results than when used independently. The study showed a significant lift in onsite engagement and an increase in online and offline purchasing by consumers who are exposed to integrated campaigns that employ both types of online advertising.”

And from Google…or formerly Google

“Grant MacFarlane, head of search at Havas agency Media Contacts and previously head of client services at Google, agrees. “Search often takes the glory for the purchase, as a lot of agencies can’t track back to see where someone saw a banner ad that got them searching in the first place,” he says. “We’ve seen a 15% or 20% uplift when you run display and search in tandem.”

There are tools that help clients track how their different ad types are performing in relation to one another.  Yahoo provides an online advertising tracking solution called “Full Analytics” that measures how that display ad was driving additional conversions to your other ad channels.

So the next time you’re planning your online ad campaign, try to do a mix of both. Measure your ROIs and see whether two, is better than one.

 
 

Online Form Labels: Left, Right or Top Aligned?

Posted by Gul Khan in Measurement on September 13th, 2008

Form Label Alignment: Left, Centre or Top?

In the past, as a general rule, I have always suggested that the form labels be right aligned. There were 2 main reasons for this recommendation:

  1. We noticed that, with the left-aligned labels, some of the users ended up entering the data in the wrong input field (e.g. the input field one line below) due to the long distance between the label and the input field.
  2. From what we observed, forms with right-align labels were perceived to be visually less cluttered (this was great for forms that would only be used once or very infrequently).

However, recently I have come across some excellent work done by Luke Wroblewski and Matteo Penzo that took my understanding of label alignment to another level. Here’s a summary of what they have reported:

LEFT-ALIGNED LABELS

Form Label: Left-Aligned (Gmail)

  • Easy to scan through the labels. This is especially useful if the form is long and it is likely to be filled by users unfamiliar with the form (they are more likely to scan through the form to determine what information is required of them).
  • Long forms, with many optional input fields, can possibly benefit from left-aligned labels, because users can easily scan what is required to be filled.
  • Requires less vertical space.
  • Slowest completion rates (this may or may not be a bad thing…it really depends on your objectives).
  • Poor association (i.e. long distance) between the label and its corresponding field. This can result in data entry errors. The effect is worsened when some labels are very long.
  • Not flexible enough to deal with large changes in label lengths.

RIGHT-ALIGNED LABELS

Form Label: Right-Align

  • Good association between the label and the corresponding field. Among other things, the short distance tends to result in reduced number of data entry errors.
  • Requires less vertical space.
  • Fast completion times (but still slower than top aligned forms).
  • Reduced readability of labels, especially with long forms. This becomes an issue for new users who are unfamiliar with the form. The left rag of the text makes it difficult to scan though the labels to find out what is required of them to complete the form. However, short forms like the one on Facebook, are unlikely to suffer from this problem. Yahoo also shows how to overcome this problem by breaking up the form into clear sections.
  • Not flexible enough to deal with large changes in label lengths.

TOP ALIGNED LABELS

Form Label: Top Aligned

  • Fastest completion times (because users could capture both the label and the input field with a single eye movement).
  • Works well when people are familiar with the data being collected.
  • Good for localization. Works well if the form is being translated simply because of the space it allows for variations in label lengths (e.g. German and French words can be twice as long as English).
  • Requires the most vertical space.
  • Requires sufficient spacing in-between lines.

Which is the best format? Well, that depends on the context.

  • How familiar are your users with the form or the individual labels?
  • Is the form going to be used frequently? If ‘yes’ then efficiency becomes an important factor. In that case, a top-aligned form could be the way to go. If it is a registration form, which your users will probably only fill in once, then you could consider left-aligned labels (if it is a long form) but personally, I feel a better option would be to use right-aligned labels and break it up into very visually clear sections.
  • If localization is an issue and you need to translate the form into many different languages, then top aligned labels gives you the most flexibility.
  • Have you got vertical space limitations? Then you may want to avoid top aligned forms.

The best advice I can give is to understand your context and weigh your options. The points above will give a sense of what will work. Use the information to try out different designs. The next step would be to test out different design version by conducting AB Split Testing, Multivariate Testing or Taguchi Testing methods. Of course there are many other testing methods out there. In my opinion, however, these tests are simplest to execute. It is always a good idea to get actual usage data to back up your design ideas. Many times, the results can surprise and give you new insights about what design works best for your users.

 
 

Using the KANO Model to Improve Customer Satisfaction

Posted by Gul Khan in Measurement on September 4th, 2008

Handling customer requirements for web projects can be a very daunting task. How do we break down your customer needs into a meaningful implementation roadmap? What are the basic features and functions that we must have? What are the “wow” factors? On top of that, how do we reconcile customer needs with business requirements?

I’ve been using the KANO model for some time now with great effect for several of our web development projects. So far, it has worked better than all expectations. Basically, it helps us to break down the customer requirements (derived by conducting various types user research) into the Basic, Performance and “Wow” components and allows us to adopt a tactical approach to website re-design roadmaps. To those who are familiar with the KANO model will immediately understand what I’m driving at. To the rest who are not familiar with this model, here’s a quick explanation:

8-Minute KANO Model Tutorial

The KANO model essentially breaks features into three separate categories:

  • The “must have” features (i.e. the basic requirements)
  • The “performance” features
  • And lastly, the “excitement” or “wow” features

How do these three different types of features affect customer satisfaction?

  • Basic: Customers get pretty annoyed when this type of feature is missing but they are neutral when it is available. For example, a hotel room with no toilet paper.
  • Performance: Customer satisfaction increases as this type of feature is better implemented; example the fuel efficiency of a car.
  • Excitement or Wow features: Customers won’t be too disappointed if these are missing but having these features can result in great customer satisfaction and it can differentiate you from your competitors.

If you want to learn more, then c2c Solutions have a great 8-Minute KANO Model Tutorial. I highly recommend it to anyone who wants to learn more about this model.

Read the rest of this entry »

 
 

Creating Usable Websites Workshop (August 20, 2008)

Posted by Gul Khan in Measurement on August 28th, 2008

Another month and another workshop! The great thing about these workshops is that every month I get to meet new people from different walks of life. It really helps me to see things from many different perspectives.

I have been doing these workshops for the government for about 6 years now and one trend that I have noticed is that the participants are becoming less and less…”gahmen”…if you know what I mean. Of course, I meant that in the nicest possible way :)

This last workshop was a real blast. You guys really made it an enjoyable session. Thanks to all the participants, starting with Xinying, Say Fern, Adelyn, Allan, Cheng Kok, Rosita , Mei Hui, Rohaizan, Margaret, Soo Chin, Amirrudin and Jian Ping. Thanks again and do keep in touch!

 
 

Creating Usable Websites (June & July 2008)

Posted by Gul Khan in Measurement, News on July 23rd, 2008

Finally, I’m getting around to posting the pictures from my Creating Usable Website training workshops which I conduct at PR Academy.

Starting first with my workshop on the 15th of July:

15th July 2008

Ranjini, Raihidaya, Amanda, Melanie, that’s me, Chiew Leng, Margaret, Ion, June, Ismiati and Joann.

Here’s the picture for the June 25th Workshop:

25th June 2008

Teow Gay, Christina, Anna, Wendy, Vanesa, Shumin, Seck Leng, Wee Chyn, Shamsiah, Kelvin, Kian Pin, Peggy.

Thank you all for your participation in the workshop. I enjoyed both sessions thoroughly and I hope I get a chance to catch up with each and every one of you in the near future :)

 
 

GoogleFIGHT! “Website” vs “Web Site”

Posted by Gul Khan in Measurement on July 7th, 2008

While I was writing my previous post about the PetrolWatch Customer Satisfaction Survey results, the text editor kept prompting me that the spelling was “web site” and not “website”. My first reaction was to get on to Google Trends or Googlefight and find out how the rest of the world was spelling it. Here’s how it works:

STEP 1: Go to www.google.com/trends and key in “website, web site” into the text box

Google Trends Home Page

Step 2: And the winner is…

Google Trends: Results for \

If you look at the two lines above the time line, you will see more people search for “website” compared with “web site”. End of argument. I’ll just add “website” to my dictionary. On occasion, we do use this technique in our website development work, especially in areas where we need fast customer insights about which is the most appropriate label to use for navigation, page titles, links and so on.

Why is this important?

Well, if you care about your search engine rankings, you do want to make sure that your keywords match the most popular keywords that your customers are using. For those of you who have had experience designing and managing websites, you will know that finding the right label and wording can sometimes be a nightmare task:

  • “Should we call it “cheap flights” or “low fares”? (This example was used by Gerry McGovern in one of his web seminars which I had the pleasure of attending)
  • “Point and shoot camera” or “point and click camera”?

In the past, we were in the dark about what is the correct label to use. Now we have free tools at our disposal that give us instant insights.

Read the rest of this entry »

 
 

PetrolWatch Survey Results (May-June 2008)

Posted by Gul Khan in Measurement on June 26th, 2008

xblog_001_findings.jpg

“PetrolWatch gives drivers an easy access on the updates of fuel prices for all the 4 fuel stations. It also provides a clear guide for available discount to help drivers made better decision to save money on fuel.”

- Comment from PetrolWatch Visitor

PetrolWatch.com.sg was launched in 2005 and the last design revamp was way back in 2006. It’s due a design facelift and so, we ran an online survey from late May to early June 2008. In that short time we managed to collect over 405 responses. Overall, PW scored 6/10 in the customer satisfaction score and successful task completion rates were around 72%. Not bad for a pro-bono project by Comwerks Interactive. To everyone who participated, the team would like to say a big “thank you”! Your comments were really a big help. It gave the team a huge lift. They are currently working on the new design right now and for all the PetrolWatchers out there…

Here’s a Sneak Preview of the New Design

PetrolWatch New HomePahe

Your Comments Have Helped Shape The Design

Here are some of the more interesting comments from the survey, good and bad. I’ve given the Editor of Petrolwatch the chance to reply ;)

quote_1.jpg

quote_2.jpg

PetrolWatch: Unlike before, it is getting increasingly harder to receive tip-offs for impending petrol price spikes/dips. Previously, we also used to be on the fuel companies’ media distribution list. However, with prices moving only northwards, things have suddenly been less chummy. Go figure!

quote_3.jpg

PetrolWatch: Actually, we do indicate the time of price changes accordingly for each brand in our NEWS section. Most of the time, all 4 brands move between 2 - 12 hrs of each other. If you look hard enough, you’ll be able to suss out the usual suspects.

quote_4.jpg

PetrolWatch: Yea, its not easy calling 3000 over carparks everyday y’know? Haha! But seriously, it became quite a logistical challenge to update the prices accurately. I know we disappointed a lot of folks out there but we’re working on getting the system smarter, and more efficient. This is one of the main reasons why we’re gonna be adopting a more “wiki” styled approach for the next version. Watch out for it!

quote_5a.jpg

quote_5b.jpg PetrolWatch: Now we’re talking..*smug smile*

quote_6.jpg

PetrolWatch: Your wish is granted! The new design will have this!

quote_7.jpg PetrolWatch: Sounds like a good idea. We’ll definitely consider putting that in as an enhancement once the new site is up and running.

And lastly…

quote_8.jpg PetrolWatch: No need la. The new site will be FREE to all. So no more subscriptions fees whatsoever!

Top Ten Things That People Wanted to See in PetrolWatch

  1. Petrol price comparison in one glance (including trends and credit card discounts)
  2. Location of petrol stations
  3. COE prices and price change over time
  4. Car park information (cheapest car park)
  5. Tips & tricks for drivers
  6. Car reviews & car accessory info
  7. Car wash info (places offering cheap car wash & vacuum services)
  8. More lively forums and blog (I want to interact with other drivers)
  9. Traffic info
  10. No more subscription plans! Make it free for all!

Customer Satisfaction Score (60%)

xblog_pw_score_june2008.png

Overall, Comwerks scored 60 out of 100 in the customer satisfaction scores. We asked people to rate the website on a scale of 1 to 10. This very useful for us because it tells us where we stand right now. We’ll be measuring again after we launch the new design so that we can capture the “BEFORE” & “AFTER” results.

The Major Problem Areas

xblog_pw_problem_june2008.png

The next thing we needed to figure out was to identify the major problem areas. The biggest gripes were hard-to-find information, incomplete content and the lack of a search feature.

  • The main reason why information was hard to find was because of the fact that you had to sign up to use most of the services in PetrolWatch.
  • The second reason was that it was not easy to view the information at a glance. And some really information were practically hidden (e.g. Price Guru). The new designs solves all this.
  • Another problem was the fact that the information was out-of-date. The guys at PetrolWatch know this and they have sorted things out to keep things ticking along nicely.

72% of Participants Could Complete the Their Tasks On the Website

xblog_pw_task_june2008.png

Overall, the top 3 tasks that people wanted to perform on on PetrolWatch are:

  1. Check current petrol prices and price trends
  2. Check for credit card discounts
  3. Look for car & petrol related information (e.g. tips on fuel efficiency, car accessories, news)

We asked the visitors if they were able to complete the task that they set out to do and 72% said “YES” while 28% said “NO”. This showed the PetrolWatch team that they really needed to focus on getting this basic information out as easily as possible, so that people can quickly compare and get a grip on prices and discounts. It also pointed to a navigation re-design.

Look out for the new PetrolWatch in end July!

Once again, to everyone who participated in the survey, your comments really helped us a lot. The PetrolWatch team are working hard to get the new design launched. We hope you like it :)